Start typing to search.
Ranks providers by their recent typical output speed in tokens per second. Providers without enough recent data remain visible after ranked rows.
Groq currently ranks first at 547.0 tok/s via groq.
| Rank | Provider / route | Output speed | Speed measurements | Endpoint details |
|---|---|---|---|---|
| #1 | Provider / route Groq groq | Output speed547.0 tok/s | Speed measurements
| Endpoint details
|
| #2 | Provider / route Weights & Biases wandb/fp4 | Output speed156.0 tok/s | Speed measurements
| Endpoint details
|
| #3 | Provider / route Google Vertex google-vertex/us-central1 | Output speed144.0 tok/s | Speed measurements
| Endpoint details
|
| #4 | Provider / route Fireworks fireworks | Output speed126.0 tok/s | Speed measurements
| Endpoint details
|
| #5 | Provider / route DeepInfra deepinfra/bf16 | Output speed107.0 tok/s | Speed measurements
| Endpoint details
|
| #6 | Provider / route Together together | Output speed104.5 tok/s | Speed measurements
| Endpoint details
|
| #7 | Provider / route Parasail parasail/fp4 | Output speed98.0 tok/s | Speed measurements
| Endpoint details
|
| #8 | Provider / route Amazon Bedrock amazon-bedrock | Output speed60.0 tok/s | Speed measurements
| Endpoint details
|
| #9 | Provider / route Phala phala | Output speed56.0 tok/s | Speed measurements
| Endpoint details
|
| #10 | Provider / route NovitaAI novita/fp4 | Output speed27.0 tok/s | Speed measurements
| Endpoint details
|
| #11 | Provider / route DekaLLM dekallm/bf16 | Output speed17.0 tok/s | Speed measurements
| Endpoint details
|
| #12 | Provider / route SiliconFlow siliconflow/fp8 | Output speed11.0 tok/s | Speed measurements
| Endpoint details
|
| — | Provider / route Amazon Bedrock amazon-bedrock/eu-west-1 | Output speedInsufficient recent data | Speed measurements
| Endpoint details
|
Endpoint data fetched .
gpt-oss-20b endpoint guide
12 of 13 gpt-oss-20b endpoints currently have the published data required for this ranking. Groq leads at 547.0 tok/s via groq. The table keeps unranked routes visible so missing measurements do not look like missing provider availability.
The output-speed ranking compares recent median generated tokens per second. It favors routes that complete long generations quickly, while first-token latency remains a separate measure of how responsive the endpoint feels at the beginning.
Weights & Biases currently ranks second at 156.0 tok/s. Compare that gap with input and output price, response-start time, output speed, 30-minute uptime, context length, and the exact route before deciding whether first place is meaningful for your workload.
gpt-oss-20b has 13 published provider options, and performance data is matched to each exact OpenRouter routing tag. Different quantization, context, regional deployment, or provider configuration can change price and behavior even when the underlying model name is identical.
